What a standardised extract actually means

Standardised means measured. A standardised extract is one in which a named compound, or a named group of compounds, is present at a stated percentage, and that percentage is confirmed by laboratory assay on each batch rather than assumed from the plant it came from.

It is the difference between an ingredient that is the same thing every time and an ingredient that merely has the same name every time.

The problem it solves

Saffron is an agricultural product. The stigmas come off flowers grown in a particular field, in a particular season, harvested by a particular crew and dried in a particular way, and the compounds inside them vary with all of that.

Grind those stigmas and you have an honest saffron powder that is a slightly different ingredient from one harvest to the next. In a kitchen that is part of the charm. It becomes a problem the moment anyone wants to know whether a published research finding applies to the material in front of them, because a trial tested one specific batch of one specific preparation.

Standardising fixes the target. The manufacturer extracts to a specification, measures the result, and the batch either meets the number or it does not ship.

What saffron is standardised to

The saffron extract with the most published placebo-controlled trials behind it, affron, is standardised to more than 3.5 per cent Lepticrosalides. That is a patented composite measure of saffron's own bioactive compounds rather than a single chemical, which is unusual enough to be worth unpacking.

The split has been published. In the methods section of a 2026 trial the researchers record it as 3.48 per cent crocins plus 0.03 per cent safranal. (Lopresti and Smith, Frontiers in Nutrition, 2026;13:1838513.) Crocins are the red pigments. Safranal is the aroma compound, and there is very much less of it.

The patents behind the marker set upper bounds on what the process can deliver, at 6.00 per cent total crocins and 0.35 per cent safranal. (Mena-Garcia et al., Foods, 2023;12:1634, citing patents US2019099464A1 and ES2573542B1.) The figure that appears on the specification is not that ceiling. It is the floor: more than 3.5 per cent, on every batch.

That distinction is worth carrying to any label. A more-than figure is a minimum each batch has to clear. An average is not, and a typical value is not either.

What gets assayed, and how often

On a saffron extract specification the marker assay is not the only per-batch test. Identity, the extract's own chromatographic fingerprint, the marker percentage, bulk density, ash, loss on drying, particle size and the routine microbial counts are all listed as analysed on every production batch.

The extraction solvent is on the same sheet and is worth reading. For this extract it is water, and nothing else.

The evidence that the number holds

Standardisation is a promise about consistency, and consistency is checkable. For this extract it has been checked and published.

Mena-Garcia and colleagues analysed 88 samples taken from different manufactured batches. Across all 88 the crocin and picrocrocin profiles were reproducible, with the ratio of trans to cis crocin isomers ranging from 3.91 to 7.26 across the whole set. The authors concluded that the manufacturing procedure behind it was robust. (Foods, 2023;12:1634.)

They also tested how long the profile lasts. Two samples were stored at 25 degrees Celsius and 60 per cent relative humidity and re-analysed at 18 months and again at 36 months. At 36 months the individual crocin isomers had shifted by a relative standard deviation of less than 3.2 per cent, with a single exception among the minor isomers.

That is the analytical basis for a 36-month shelf life. It is a measured result rather than an estimate, which is the whole idea.

Three questions that separate an extract from a powder

• What is the marker? It should have a name. Standardised on its own, with nothing after it, is not a specification.
• What is the percentage, and is it a minimum? Look for more-than rather than approximately.
• Is it assayed on every batch? A single historic analysis describes one batch that has long since been sold.

None of this is difficult for a manufacturer who is already doing the work, which is why the answers arrive quickly when the work is being done.

Studies conducted on individual ingredients, not on any finished product. Individual results may vary.